The cytolytic enzyme produced by bacillus carotovorus and certain other soft rot bacteria

ifgiPs ' ^ 111 'fe&v • '£;*% .^£';^‘*'--' if ip|| V: ' -'" fg

|f|!^

II grayKgSf aKi^ '-F&*$^'hif4 f '

BBs|B8w^gMpww8NpgNfe|^^-:-:^^('^i

.

mi n

«&■!«««
j^j.4»KH5«,-A«..- **-« j . .-* a-.. jjjjg >V. •,

jL^jKS** < * :>.v:

g|pd|Igl|2

■■H

Slip

-fc

CYTOLYTIC ENZYME PRODUCED BY
lCILLUS carotovorus and cer-

&4

AIN OTHER SOFT ROT BACTERI

m$.

■y« f

mmM

wBm

;- :> m

W;S^':

>r of Philosophy by Lewis Ralph Jones

^'vV

\0 \tv\B

Abdruck aus dem

Centralblattf. Bakteriologie, Parasitenkunde u. Infektionskrankheiten.

II. Abteilunff.

•Herausgeg. von Prof. Dr. O. Uhlworm in Berlin. — Yerlag von Gustav Fischer in J ena.

XIV. Band. 1905. No. 9/10.

Nachdruck verboten.

The cytolytic enzyme produced by Bacillus carotovorus
and certain other soft rot bacteria.

By Li Ri Jones,

Professor of Botany in the University of Vermont. O

An account of our studies upon the soft rot of the carrot caused
by Bacillus carotovorus was published in this journal three years
ago1)- The following statements occur therein.

Wahrscheinlich werden weitere Studien iiber die bakterielle Weich-
faule’von Vegetabilien zu der Entdeckung vieler ahnlicher Organismen
ftihren, vomJl^jeder fahig ist, eine groBere Reihe von Wirtpflanzen
^zu befallen^^Wst sicherlich von Wichtigkeit, sowohl vom Gesichts-
^punkt der abstrakten Bakteriologie, als von der Pflanzenpathologie aus,

^daB sie vollstandig bekannt werden Die mikroskopische Unter-

suchung des faulenden Mohrengewebes hat gezeigt, daB der Organismus
in die Intercellularraume eindringt und sich darin mit ungeheurer
Schnelligkeit vermehrt. Die mittleren Lamellen der ausliegenden Zellen
erscheinen erweicht oder zerstort durch die Ausscheidung von Extrakten
der Bakterien, denn in ergriffenen Geweben findet Isolierung der Zellen
statt. Das Protoplasma in solchen isolierten Zellen ist zusammengefallen,
aber man hat keine Bakterien in den Zellen von frisch desorganisiertem
Gewebe bemerkt. Diese Zerstorung der Intercellularsubstanz ruhrt
wahrscheinlich von einem Enzym von der Natur der Cytase her . . .

The correctness of the first suggestion has already received ample
proof both from further unpublished studies on soft rot organisms in
our own laboratory and from several publications dealing with these,

1) Centralbl. f. Bakt. Abt. II. Bd. VII. 1901. p. 12 and 61. For a more detailed
account of the same see Report Vermont Experiment Station. XIII. 1900. p. 299.

The present publication deals with a direct continuation of the earlier studies and
the reader is referred to the former publication for details relative to the source and
character of the organism, culture methods, etc.

Most of the results here given were embodied in a paper read Dec. 1902 before
the Society for Plant Physiology and Morphology, Washington, D. C. Sincethe work
was undertaken papers have been published by Spieckermann, Van Hall and
Potter dealing with some of the same questions. Reference to these is made later in
Xour text. Since our own studies were largely completed before these came to our
hands it follows that the conclusions where they are identical are the more firmly
established. Where our results differ from any of theirs we have carefully repeated
the work and have satisfied ourselves of the correctness of our position.

This work forms one part of a comparative study of various soft-rot bacteria
which has been conducted in cooperation with Messrs. H. A. Harding and F. C.
'I'Stewart of the Geneva, N. Y. Experiment Station and W. J. Morse of our institu-
tion. The full details are reserved for a joint publication from these two stations to
be made soon.

Much of the study was carried on by the author while in residence at the Uni-
versity of Michigan and he gratefully acknowledges advice from Prof. F. C. New-
combe and also excellent help in many of the details from two student assistants,
Messrs. H. D. Bone and L. P. Sprague in the work done in his own laboratory.

„ _ ™ 17

258

L. R. Jones

including those of Townsend1), Harding and Stewart2) and
Harrison3) in America and Spieckermann4), Van Hall5) and
Appel6) in Europe.

A fuller understanding of the wall-dissolving or other enzymes pro-
duced by these bacteria is desirable for two reasons. First, because
the parasitism in each case seems dependent upon enzyme production;
second, because these organism are so similar as to make the question
of their specific relationship an extremely complex one which must be
settled largely by appeal to physiological characters and it would seem
that enzyme production might prove of value for such differential pur-
poses.

It is also of general interest to learn whether or not these bacte-
rial cytolysts are identical with the cytolytic enzymes known to be deve-
loped by certain fungi, germinating seeds and pollen tubes. This latter
question has been clearly raised by Green7), who entertains the belief
that these so-called “cytases” include two distinct kinds of enzyme.

The action of the organism upon the tissues of the

host plant.

This was discussed in our earlier publications. Our later studies
have confirmed the earlier conclusions that the action leads to the death
of the cells and to the solution of the middle lamellae of the paren-
chymatous tissues and leaves an undissolved residual wall through
which the organisms do not pass. This strictly intercellular
invasion is in accord with observations made by Spieckermann and
Van Hall upon their similar organism and also by Appel but is
opposed to that of Potter8) who reports penetration of the cell
cavities.

As seen by the naked eye the invasion is marked by a rapid soft
rot of the parenchymatous tissues, accompanied by a water-logged ap-
pearance and often with discoloration. The boundery between invaded
and sound tissues is always clearly defined. Cuticularized and lignified
tissues are unaffected.

Under the microscope the collapse and evident death af the proto-
plasm is seen to occur promptly, and co-incidentally with the following
changes in the cell-membrane. There is a loss of refractiveness in the

1) Townsend, C. 0., An unpublished paper on a bacterial soft rot of the Calla.
Read before the Society for Plant Physiology and Morphology, New York. Dec. 1901);
A soft rot of Calla Lily. (U. S. Dept. Agric. Bureau Plant Industry. Bulletin 60.
June 30, 1904.)

2) Harding, H. A. and Stewart, F. C. , A bacterial soft-rot of certain cruci-
ferous plants and Amorphophallus simlense. (Science. N. S. Vol. XVI. 1902. p. 314.

3) Harrison, F. C., Preliminary report on a new organism producing rot
in Cauliflower and allied plants. (Science. N. S. Vol. XVI. 1902. p. 152.)

4) Spieckermann, A., Beitrag zur Kenntnis der bakteriellen Wundfaulnis der
Kulturpflanzen. (Landw. Jahrb. von Thiel. 1902. p. 155.)

5) Van Hall, C. J. J., Das Faulen der jungen Schofilinge und Rhizome von
Iris florentina und Iris germanica. (Zeitschr. f. Pflanzenkr. Bd. XIII. 1903. p. 129.)

6) Appel, Otto, Untersuchungen uber die Schwarzbeinigkeit und die durch
Bakterien hervorgerufene Knollenfaule der Kartoffel. (Arb. a. d. biol. Abt. f. Landw.
u. Forstw. am Kais. Gesundheitsamt. Bd. III. 1903. p. 364.)

7) Green, J. Reynolds, The soluble ferments and fermentation. 2. edition.
1901. p. 105.

8) Potter, M. C., On the parasitism of Pseudomonas destructans. (Proc. Roy.
Soc. London. Vol. LXX. 1902. p. 393.)

The cytolytic enzyme produced by Bacillus carotovorus etc.

259

inner lamellae of the membrane (i. e. the portion lying between middle
lamella and cell cavity) which is accompanied by its strong swelling,
often to twice the original thickness, and the appearance of delicate
lamination.

The middle lamella scarcely changes its refractive character and
is therefore brought out in sharp contrast with the inner lamellae.
Neither does it swell, but instead it soon melts away, disappearing
first in its thinner parts while the thickened masses at the angles of
the cells persist longest but ultimately dissolve also.

Before the latter are all dissolved the cells lose their cohesion, i. e.
the tissues are fully rotten. Thin sections of carrot or turnip tissue
immersed in culture broths or the enzyme solutions (to be described
later) generally pass through the above changes in from ten minutes to
one hour’s time. The disappearance of the middle lamella marks the
completion of all visible action. The swollen remains of the inner la-
mellae are not further acted upon even after many days and give the
characteristic cellulose reaction with chlor-zinc-iodide.

In order to follow some of these details more carefully, blocks of
carrot and of turnip tissue in process of invasion by the organism were
fixed in absolute alcohol, imbedded, sectioned and stained in various ways.
These have shown that the swelling of the walls and the partial solu-
tion of the middle lamella occurs some distance, often ten cells, in ad-
vance of the actual invasion by the organism. Such sections have also
shown that the middle lamella like the inner lamella shows a laminated
structure when partially dissolved.

Briefly, then, the action consists primarily in the rapid and com-
plete solution of the middle lamella, or pectic layer, of the wall. This
is accompanied by the swelling of the inner lamellae, or hemicellulose
layers and the revelation of a laminated structure in them which is
evidently due to the occurrence of soluble pectic layers alternating with
the insoluble hemicellulose layers. The hemicellulose residuum although
evidently softened still serves to debar the organisms from entrance into
the cell cavities, no organisms having been observed in the interior of
either the cells or the vessels of unmutilated tissues, even in advanced
stages of decomposition and when the intercellular spaces were crowded
with the organisms.

Careful tests have been made at various times and with cultures
in different media for evidence of diastase (amylase) but none has been
found.

This action upon the cell membranes was evidently of an enzymic
nature. The chief task we set before ourselves was to isolate this en-
zyme from the living organism and study its characteristics. Five me-
thods have been tried for accomplishing this separation : heat, filtration,
the use of germicides, diffusion, precipitation with alcohol.

The first three of these methods have involved in all cases alike
the following procedures: The cultivation of the organism in beef or
vegetable broths for several days ; treatment of such broth cultures
aiming to kill or remove the bacilli without injury to the enzyme; de-
termination by trial transfers of the sterility of the broth so treated;
in case sterility was proved, the testing of its cytolytic activity by im-
mersion in it, under sterile conditions, of vegetable tissues. In all
cases control trials were made with the uninoculated broths or solu-

17*

260

L. R. Jones,

tions of the chemicals under trial to preclude the possibility of cytolytic
action by these alone.

We will now very briefly summarize the results obtained by each
of these five methods, leaving further details for elucidation in our
later publication.

Separation of the enzyme from the living bacteria

by heating.

Broth cultures of Bacillus carotovorus are sterilized by ten
minutes exposure to a temperature of 51 0 C. Previous students x) of
cytolytic enzymes have found them to be destroyed at about 60° C. It
seemed probable, therefore, that by heating broth cultures to some in-
termediate temperature one might kill the bacteria and leave the enzyme
active. This was attempted in a series of tube cultures, 10 ccm each,
heated by immersion in a water bath. The results were uniform and
satisfactory. Cultures so exposed at each 54, 55, 58, 60, 62, 63, 64,
65, 68, 73° were sterilized. Subsequent trial showed that cytolytic acti-
vity remained in the broths so heated until 62° was passed when it
was lost. There was however some weakening of the cytolytic action
by such heating at 54° and much more at 58°. Slight activity per-
sisted in broths heated at 60°, and a trace at 62°. Above this there
was none. In brief then there was slight inhibition of activity as a
result of such heating for ten minutes at 54 °, decided inhibition at 58 °,
and total inhibition at 63°. A further discussion of temperature rela-
tions occurs later in this article.

Separation of the enzyme from the bacteria by filtration.

The Paste ur- Cham b er lan d filter was used and the usual pre-
cautions taken to prove the continued sterility of the filtrate. In all
our trials cytolytic activity has been found in the sterile filtrate, and
the action upon vegetable tissues was indistinguishable in character from
that occurring in the presence of the living organism. The rate of
this cytolytic action in the filtrate was however decidedly less than that
in the original broth. In some cases this loss was estimated at 90 per
cent, but in none was it complete.

These results may profitably be compared with those of Potter1 2),
Laurent3) and Van Hall4), each of whom found similar bacterial
filtrates to retain their cytolytic activity, whereas Spieckermann5)
found the sterile filtrate from cultures of his kalerot organism to have
entirely lost its enzymic activity. If one seeks an explanation of this
loss through filtration it may be attributed to either of two things,
first, the retention of only that portion of the enzyme which is still
closely associated with the organisms, i. e. within the cells or upon
their surfaces; or, second, the further removal of enzyme which has

1) Cf. Green, loc. cit. p. 98.

2) Potter, M. C. , On a bacterial disease of the turnip. (Proc. Royal Soc. Lon-
don. Vol. LXVII. 1900. p. 448.)

3) Laurent, E. , Recherches exp^rimen tales sur les maladies des plantes. (Ann.
Inst. Past. T. XIII. 1899. p. 1.)

4) Van Hall, C. J. J. , Bacillus subtilis und B. vulgatus als Pflanzenparasiten.
(Centralbl. f. Bakt. Abt. II. Bd. IX. 1902. p. 642.)

5) Spieckermann, loc. cit.

The cytolytic enzyme produced by Bacillus carotovorus etc.

261

already passed from the organisms into solution in the broth. The
latter theory is favored by Friedenreich’s observations which also
aid in reconciling the apparent discrepancies between Spiecker-
mann’s results and those obtained by ourselves and others. Frie-
denreich1) working with cheese extracts found that porcelain filters
might retain as high as 90 percent of the protein matter which was in
solution and that the amount so retained was much greater with bou-
gies which had been used repeatedly than with new ones. In some of
his trials the milk enzyme galactose was entirely removed by filtration.

Separation of the enzyme from the living bacteria

by germicides.

In our attempts to accomplish this use has been made of chloro-
form, phenol, thymol and formalin. Trial additions of various amounts
of each of these have been made to broth cultures of the carrot-rot
bacillus and the effects noted upon the life of the organism and the
activity of the enzyme.

Formalin. Merck’s formalin was used in all cases. It has
been found that both the bacillus and the enzyme are extremely sensi-
tive to this chemical. Since, however, the bacillus is slightly more so,
it is possible, by carefully regulating the amount, to sterilize the broth
and leave the enzyme active. The addition of 0,1 percent of formalin
to beef broth tube cultures of B. carotovorus sterilizes, providing
there is thorough agitation. More formalin, viz., 0,2 percent or
more, may be necessary to sterilize if the agitation is less thorough.
Complete inhibition of the enzymic (cytolytic) activity in such culture
broths occurs where 0,6 percent or more of formalin has been added.
Amounts as low as 0,3 percent retard the action to a marked degree
and retardation was perceptible when even 0,06 percent was used, al-
though this is not enough to sterilize with certainty. Trials where the
formalin was added to aqueous solutions of the enzyme-containing alco-
holic precipitates (see discussion later) gave results in accord with the
above, viz., a slight but appreciable retardation from additions of 0,5 per-
cent and almost complete inhibition where 0,5 percent was used.

In all of the above cases the determinations of enzymic activity
were made a day or more after the formalin was added to the broth.
After the completion of these trials Spieckermann’s2) report came
to hand in wich he states that 0,2 percent of formalin sterilized cul-
tures of his kale-rot organism and did not inhibit the action of the
enzyme, at least for several hours. This statement led us to repeat
our observation with the carrot-rot bacillus that we might learn the
relative rapidity of the injurious action of formalin upon each the or-
ganism and its cytolytic enzyme. The details must await our later
publications. The action was as Spieckermann had observed, more
rapid upon the bacillus than upon the enzyme. Thus, 0,2 per cent of
formalin caused appreciable injury to the organism at the end of about
three hours, as shown by delay in the growth when transfers were made
from such treated broths. The lessening of cytolytic activity in such

1) v. Friedenreich, E. , Landw. Jahrb. der Schweiz. Bd. XIII. 1899. p. 169.
Ann. Agr. Suisse. T. I. 1900. p. 77.

2) Spieckerm ann , loc. cit. p. 166—167.

262

L. K. J ones ,

broths became apparent only after a longer delay, viz., nine hours more
or less.

These results showed that formalin cannot be employed as an agent
for the purposes we wished since if one makes the studies within a few
hours after the additions of the formalin sterility of the culture broths
is not insured; if, on the other hand, one waits much longer the acti-
vity of the enzyme may be impaired.

Bliss and Novy1 2) have shown that fibrin which has been acted
upon for a short time by formalin resists thereafter the action of proteo-
lytic enzymes. Their observations led us to wonder whether any part
of the retarding influence of formalin observed in our experiments
might be due to similar action of the formalin upon the cell walls of
the vegetable tissues rather than upon the enzyme. Trials showed that
this is not the case, since vegetable tissues immersed in formalin and
later washed in water were decomposed in the presence of the enzyme
as readily as were fresh tissues.

Phenol. Numerous trials of this gave satisfactory results. A
crystal about one-half the size of a pea added to a 10 ccm broth cul-
ture , i. e. making approximately 0,5 percent solution , well agitated,
has in all cases secured sterility and caused no appreciable lessening
of the activity of the enzyme.

Thymol. This has given variable results, sometimes sterilizing
and sometimes not. These irregularities are, we believe, due to the
slight solubility and slow diffusibility of thymol in the broth. It was
found that powdered thymol floats on the surface of the broth even
though well shaken. This has sterilized the upper portions of broth
tubes, when added in excess, while living organism persisted indefini-
tely, at least for many days, in the depths of such tubes, providing
they were not repeatedly agitated. Thorough agitation did, however,
secure sterilization where 0,2 percent or more of thymol was used. In
no case did thymol cause appreciable lessening of the activity of the
cytolytic enzyme.

Chloroform. Since this is the agent usually employed for the
inhibition of bacterial growths in the study of enzymes, careful trials
were made to determine its relation to the activities of both enzyme
and bacillus. Here again numerous details must await our later publi-
cation. The general results were as follows: The addition to beef broth
cultures (the ordinary tube cultures stoppered with cotton) of Bacil-
lus carotovorusof as large amounts of chloroform as 50 percent
did not sterilize where unaccompanied by thorough agitation, i. e. such
cultures may continue to increase in cloudiness and transfers show
living organisms for days there-after. In all trials, however, when 10 per-
cent or more of chloroform was added to broth cultures and these were
very thoroughly agitated sterility was secured. With less than
10 per cent the results were not constant. In no case has chloroform
caused appreciable lessening of the cytolytic action in such broths, nor
has it in trials where we have added it to the aqueous solution of the
enzyme-containing alcoholic precipitate to be discussed later.

A comparison of the results obtained by other investigators who
have tested these chemicals upon similar organisms shows a lack of

1) Bliss, C. L. and Novy, F. G. , Action of formaldehyde on enzymes and on

certain proteids. (Journ. Exper. Medicine. Vol. IV. 1899. p. 52.)

The cytolytic enzyme produced by Bacillus carotovorus etc.

263

conformity, especially with chloroform. Smith1) was the first to em-
phasize the fact that many bacteria are surprisingly resistent to the
germicidal action of chloroform and thymol. Potter2) did not succeed
in sterilizing cultures of his Pseudomonas campestris by the use
of formalin, thymol or chloroform. Spieckermanns3) results with
formalin have already been reviewed: he also found chloroform an ef-
fective agent for sterilizing cultures of his kale-rot organism and obser-
ved no retardation of its cytolytic enzyme thereby. Van Hall4), on
the contrary, reports that even 0,5 percent of chloroform acting only
fifteen minutes destroyed all traces of cytolytic activity in broth cultures
of his Bacillus omnivorus. These differences are not easily recon-
cilable. We are inclined to attribute the destruction of the enzyme in
Van Hall’s exceptional experience to some agency other than the
chloroform. In other cases the differences may be explained in part, if
not altogether, by differences iu manipulation, especially as to amount
of agitation. Certainly Smith’s conclusions are justified that chloro-
form cannot be relied upon implicitly as a germicidal agent, as has
heretofore been done so frequently in enzyme studies. Rightly used,
it has, however, proved the most satisfactory germicidal agent in our
studies.

Separation of the enzyme from the bacteria by diffusion.

As already described, the softening of the vegetable tissues occurs
several cells in advance of the actual point of invasion of the carrot-rot
bacillus. This is also in accord with the observations of several other
investigators of soft-rot organisms and indicates that the diffusion of
the cytolytic enzyme occurs quite rapidly. Van H a 1 l’s 5) recent ex-
periments with Bacillus omnivorus showed this to occur through
layers of sterile agar. We made similar trials with cultures of B. ca-
rotovorus and secured like results.

The method as we developed it consisted in implanting the orga-
nism in the middle of a poured plate of stiff (2 %) nutrient-broth
agar, the agar layer being about 5 mm thick. On the third day, when
the colony was about 1 cm in diameter, this sheet of agar bearing the
colony was transferred to the sterile surface of a freshly cut slice of
living turnip root, with precautions against contamination. Repeated
trials have shown a rapid rotting of the turnip tissues underlying the
colony with discoloration, death of the protoplasm and cytolytic action
exactly as occurs in the presence of the bacteria. Transfers of bits
from this underlying rotten turnip tissue have, however, in all cases
proved its sterility. The action must therefore be attributed to the
soluble and diffusible products of the bacterial colony on the superna-
tant agar.

Similar experiments in which the agar layer carrying the bacterial
colony was superimposed upon the surface of a sterile layer of gela-
tine have shown a corresponding liquefaction of the gelatine underlying

1) Smith, Erwin F. , Growth of bacteria in the presence of Chloroform and
Thymol. (Science. N. S. Vol. XIII. 1901. p. 327.)

2) Potter, M. C. , Proc. Roy. Soc. Vol. LX VII. 1900. p. 448.

3) S pieckerman n , loc. cit. p. 166.

4) Van Hall, Zeitschr. f. Pflanzenkr. Bd. XIII. 1903. p. 129—144.

5) Van Hall, Centralbl. f. Bakt. Abt. II. Bd. IX. 1902. p. 649.

264

L. R. Jones,

the colony, which again must be attributed to the diffused products,
since the liquified gelatine was sterile.

♦

Separation of the enzyme from the living bacteria by
precipitation with alcohol.

In this work the broth culture of the desired age was filtered in
one or another way (see details below), then 95 percent alcohol was
added, the flocculent precipitate collected, dried, and finally redissolved
in water as desired for study. In connection with this work several
questions arose as to details of methods, and experimental studies were
undertaken to decide these. The following is a summary of the results.

Filtration. The alcohol throws down, along with the enzyme,
the various proteid contents of the culture broths and the precipitate
also carries down the bodies of the bacteria. It is therefore plainly
desirable (previous to precipitation) to remove from the broth by filtra-
tion such of this other matter as one can without lessening the enzyme
content. Filtration through porcelain was followed by P otter in his work.
We tried this also in our earlier expermiments, but as already explained
we found the enzyme content of the filtrate so much reduced thereby
that we discontinued it. Thereafter we were content to filter simply
through paper which removes only the coarser deposits from the culture
broths. There is no objection to the employment of paper filters in
the manipulations with this enzyme, since it is inactive upon cellulose.

The most favorable strength of alcohol. B. caroto-
vorus is killed by 25 percent alcohol. Even this strength will throw
down a considerable precipitate, and trials were made to determine the
optimum per cent for securing the enzyme and also to learn whether a
purer enzyme could not be obtained by some method of fractional preci-
pitation or reprecipitation. Without here giving details we will merely
state that the precipitate secured in the presence of 80 percent
alcohol was found to include practically all obtainable by any higher or
lower strength, and also that the enzyme so secured showed a maximum of
activity. Nor was the enzyme much purified when the precipitate from
80 percent alcohol was dissolved in water and again precipitated with
alcohol.

Method of drying. The method commonly recommended for
securing enzymes is to wash the precipitate in absolute alcohol and dry
in partial vacuum over sulphuric acid. This process has given excellent
results in our trials, but equally good have been secured more expe-
ditiously by drying quickly in a gentle current of warm air, not above
40 0 C. To secure more rapid drying the precipitate should be broken
up and occasionally stirred. Spieckermann1) washed the precipitate
in absolute alcohol followed by ether, presumably to expedite the dyring.
We tried this method and found our enzyme much weakened. Absolute
alcohol followed by chloroform, on the other hand, gave an enzyme of
full activity but offered no advantages over drying directly from the
95 percent alcohol.

Strength of solution. The dry precipitate obtained from beef
broth cultures with 80 percent alcohol has averaged about 0,35 percent

1) Spieckermann, loc. cit. p. 165.

The cytolytic enzyme produced by Bacillus carotovorus etc.

265

of the weight of the broth. When this is added to water it swells and
is partially dissolved, and the solution so obtained shows cytolytic acti-
vity of exactly the same character as does the living broth cultures. Such
solutions were used much in our investigations and two important
questions presented themselves : first, the relation of the dilution of this
aqueous solution to its cytolytic activity; second, the relative enzymic
activity of such aqueous solutions of the precipitate in comparison with
that of the original broth cultures. Here again the details must await
our later report and only the general results now be given. Regarding
the first question, it has been found that the cytolytic activity increases
with the concentration of the solution. Thus where we compared 1 per-
cent, 5 percent and 10 percent additions of the precipitate to distilled
water a like amount of cytolytic action was secured in the following
periods of time, respectfully : with the 1 percent solution, in 25 minutes ;
with the 5 percent in 15 minutes; with the 10 percent in 10 minutes.
This shows that the enzymic activities of these solutions stood to each
other in the ratio of 6:10: 15. In our practice we have used the 5 per-
cent solution almost altogether.

The second question is of especial interest since it involves the
query as to whether the enzyme is injured by alcoholic precipitation.
As already stated, the cytolytic action of the solution, as shown alike
by macroscopic and by microscopic appearance, is of the same character
as that occurring in the original broth cultures. One might expect some
loss or weakening of the enzyme in the process of precipitation, desic-
cation and re-solution, even if the enzyme is not altered in its essential
composition. As a matter of fact, however, our trials have shown no
appreciable loss in this way. The precipitate, whether from beef broth
or vegetable juice cultures, when brought into an aqueous solution of
the same volume as that of the original culture has shown cytolytic
activity equalling that. Spieckermann1) has reported similar results
with his kale-rot organism.

The relation of cultural conditio ns to enzymic production.

1) The composition of the medium. The vigor of develop-
ment of the carrot - rot organism varies widely in different media, and
it is of much interest to learn whether there are corresponding variations
in enzyme production. De Bary2) suggested that the disorganization
products of the cell walls of the host plant are probably the chief source
of food of the soft-rot fungus Peziza sclerotiorum and that these
are rendered available by the wall-dissolving enzyme. Ward 3) concluded
that in the physiologically similar lily Botrytis the production of this
cytolytic enzyme is a starvation phenomenon. This idea is, moreover, in
general accord with the conclusion of Brown and Morris4) that
diastase secretion in germinating barley is excited by cellular starvation.
We were led to make various experiments to learn whether there is
any constant relation between nutrition and amount of enzyme produc-

1) Spieckermann, loc. cit. p. 166.

2) De Bary, A., Ueber einige Sklerotien und Sklerotienkrankheiten. (Bot. Zei-
tung. Bd. XLIV. 1886. p. 378.)

3) Ward, H. M., A lily disease. (Ann. of Bot. Vol. II. 1888. p. 319.)

4) Brown, H. T. and Morris, G. H., The germination of some of the Gra-
mineae. (Journ. Chem. Soc. Trans. Vol. LVII. 1890. p. 498.)

266

L. R. Jones

tion with the carrot-rot bacillus, and, especially whether the production
of this cytolytic enzyme is a starvation phenomenon. In these experi-
ments determinations have been made of the enzyme product from cul-
tures in media varying widely in nutritive elements and especially in
carbohydrate content. These have included Dunham’s peptone solution
with and without the addition of carbohydrates, nutrient beef broth with
and without the addition of carbohydrates, cooked vegetables entire and
the same filtered after cooking so as to remove the cell wall substance
and other insoluble parts, and finally living vegetable tissues.

Only the general results and conclusions can be given here. Cul-
tures in the fresh turnips and carrots made an exceedingly vigorous
development and the expressed juice from these decaying vegetables has
shown a higher enzyme content than has any other medium. On the
other hand, D un ham ’s peptone solution which is a starvation medium
for this organism has shown the least enzyme development. The pre-
sence of cane sugar favored the development of the bacillus and likewise
it increased the enzyme product. Thus, beef broth with 2 percent sugar
showed a more rapid growth of the organism and also a greater enzyme
product than the same without the sugar. The presence or absence of
cell-wall substance in the cooked vegetable media made no difference in
the enzyme production.

The conclusions reached are, first, that enzyme production is directly
proportional to vigor of development of the organism : second, that there
is no evidence that enzyme production is a starvation phenomenon, but
rather the reverse; third, we find no evidence that the products of en-
zymic activity are used as food by the organism, although our experi-
ments are insufficient to justify final conclusions in the latter point.

These conclusions are in general accord with those of R. E. Smith1)

in the case of Botrytis cinerea, since he holds that the fungus can
only develop its cytolytic enzyme in the presence of abundant food.

2) Age of culture. The enzyme product has been compared in

beef and vegetable broth cultures of various ages from one to seventeen

days. It has been found that the amount increases with the age of the

culture up to a certain point, then remains about constant. This increase
of enzyme content corresponds to the rate of development of the cul-
tures as shown by the degree of cloudiness of the broth, ceasing when
it passes its maximum. The conclusion already formulated seems again
justified that enzyme production by this organism is not a starvation
phenomenon but rather the normal accompaniment of vigorous develop-
ment under favorable nutritive conditions. Moreover, it is evident that
the enzyme is a fairly stable compound which, following its excretion
into the broth, tends to accumulate with the age of the culture.

3) Temperature. The optimum temperature for growth of this
bacillus is 28—30° C, i. e. broth tubes cloud more rapidly at this tempe-
rature than when above or below it. We expected, in the light of the
results just described, to find this temperature the optimum for enzyme
production also. It has been found, however, that cultures at laboratory
temperature, 18 — 22° C, develop a greater amount of enzyme than do
parallel cultures grown at a constant temperature of 30° C.

1) Smith, R. E., The parasitism of Botrytis cinerea. (Bot. Gaz. Vol. XXXIII.
1902. p. 421—436.)

The cytolytic enzyme produced by Bacillus carotovorus etc.

267

Relation of various conditions to the activity of the

enzyme.

This has been worked out largely with aqueous solutions of the
alcoholic precipitate, rather than with living cultures, because of the
greater convenience and closer uniformity of trials with such solutions.
The relation of strength of solution to its cytolytic activity has already
been discussed. The more important of the other matters worked out
are as follows :

1) Effect of long keeping. As already stated, the enzyme
seems to be a stable compound which persists unchanged for a long
time in the culture broths. It can also be preserved indefinitely as the
dry precipitate. Repeated trials with such after one or two months
keeping have shown no alteration in their activity, and in one case a
careful re-trial of such a precipitate two years old has shown no ap-
preciable loss from such long keeping.

2) Temperature. A series of careful comparisons has shown that
the aqueous solution of the precipitated enzyme exercises very little
cytolytic action at 2° C; the activity is good at 20°, better at 32°,
best at 40 — 45°. At 48° it is considerably inhibited, and it is wholly
checked at 51 °. Thus, action was practically twice as rapid at 42 0 as
it was at 22°, and at 32° it was about midway in activity between the
higher and the lower degrees. In these temperature relations therefore
it is similar to the cytolytic enzyme of germinating barley as recorded
by Brown and Morris ')• Holding at 49° for an hour’s time was not
injurious to the enzyme, i. e. although inhibited while at that tempe-
rature, it resumed normal activity thereafter as the temperature was
lowered. When, however, the temperature was carried to 51 0 or above
the enzyme was destroyed in such solutions.

In this connection it may be recalled that the trials, discussed
earlier in this article, showed that in the original broth cultures the
point of thermal destruction is some ten degrees higher. Similar varia-
tions in the point of thermal destruction under changed conditions have
been recorded for other enzymes1 2). Thus, invertase withstands a tempe-
rature twenty-five degrees higher when in the presence of cane sugar,
upon which it acts, than in its absence. Woods3; has shown that the
oxydizing enzymes of the maple leaf withstand a higher temperature
when in the juices of the plant than in the presence of alcohol. Think-
ing that the presence of the vegetable tissues upon which it is active
might restore to our precipitated enzyme its ability to withstand the
higher temperature, comparative trials were made at 51 0 where pieces
of carrot were immersed in the solutions during the heating process.
The enzyme was nevertheless destroyed, alike in the presence or the
absence of the vegetable tissue.

3) Effects of alkalies and acids. Our earlier studies showed
this bacillus to be a facultative parasite upon various vegetables, all of
which have an acid cell sap. In the course of its invasion the reaction
of the vegetable juice is rendered alkaline. It is of interest, as bearing
upon the question of parasitism and resistance thereto, to learn the

1) Brown and Morris, loc. cit. p. 502.

2) Green, loc. cit. p. 448.

8) Woods , A. F., The destruction of chlorophyll by oxidizing enzymes. (Centralbl.
f. Bakt. Abt. II. Bd. V. 1899. p. 745-754.)

268

L. R. J ones

relation of the reaction of the plant juices or other solvent to the acti-
vity of the enzyme. For learning this the alcoholic precipitate, which
is neutral, was used. The relative cytolytic activity of this upon carrot
tissues was determined when it was dissolved in water along with
various acids or alkalies. In all the following discussion the strength
of these is given as determined by titration with phenolphthalein against
normal solutions.

Alkali. It was found that sodium hydroxide titrating — 2,0 per
cent inhibited slightly and that this inhibition increased with increase
of alkalinity up to — 10,0 percent when it was total.

Acids. A very small addition of hydrochloric acid favored the
enzymic activity, about +0,5 percent being the optimum. The differ-
ence between this and the neutral solution was, however, but slight. When
the reaction reached +2,5 percent inhibition was practically complete.

Various organic acids were tested in like manner, including oxalic,
acetic, formic, tartaric, malic and citric. In no case did any of these
favor the activity. When the titration strength was below +0,5 per
cent they were practically without effect. Strengths of + 1,0 per cent
and above distinctly inhibited in all cases, and from +5,0 percent to
+ 10,0 percent caused total inhibition. It should, of course, be borne
in mind that this degree if acidity as expressed in titration percentage
is in all cases a mild one, amounting to 0,5 percent and less when
expressed gravimetricallv.

4) Relation of plant juices. The freshly expressed juice of
some of the most susceptible vegetables was used as a solvent for the
enzymic precipitate and these solutions compared with those in pure
water. It was found that in all cases tried these juices, which are, of
course, acid, lessened the cytolytic activity as compared with the water.
Thus, the juice of carrot, titrating +2,0 percent, and of radish, titra-
ting + 0,75 percent, both slightly retarded the enzymic action, and there
was still appreciable retardation when these juices were diluted to one-
half strength with water. The juice of ripe tomato which is more
strongly acid, +5 percent, was more decidedly inhibitory, reducing the
enzymic activity to approximately one half of that in the aqueous so-
lution, i. e., there was as complete cytolytic action at the end of fifteen
minutes in the aqueous solution as occurred in one-half hour in the
solution in tomato juice. This retardation was lessened when the tomato
juice was diluted with one-half water.

5) Relation to other bacterial products. As a result of
his observation upon the bacterial soft rot of the turnip, Potter1)
suggested that the oxalic acid produced by his organism, Pseudo-
monas destructans, may play some part in the dissolving of the
middle lamella. Our results, as given above, show that neither oxalic
nor any other of the acids tested so function with the carrot-rot orga-
nism. Indeed this produces no oxalic acid. It does, however, produce a small
amount of some unidentified organic acid, especially in media rich in sugar.
The following experiments were made to determine whether this acid or
any other product of the bacterial metabolism favors the cytolytic action.
Culture broths of various kinds in which the organism had developed
were heated to 80° C, which at the same time sterilized them and
destroyed their enzyme content. These were then used as solvents of

1) Potter, Proc. Roy. Soc. Vol. LXVII. 1900. p. 451.

The cytolytic enzyme produced by Bacillus carotovorus etc.

269

the enzymic alcoholic precipitate and the cytolytic activity of such so-
lutions was compared with solutions in pure water. More or less in-
hibition was apparent in every case where a culture broth was used as
the solvent as is shown in the following tabular summary of results.

Nature of solvent

Its reaction

Effect on enzyme

Carrot broth, 12-day culture

slightly alkaline

slight inhibition

Beef broth, 7-day culture

11 11

more inhibition

Dunham solution, 16-day culture

Dunham solution plus 2 per cent cane sugar,
16- day culture

neutral

much inhibition

acid

greatest inhibition

There is no evidence here that the acid or other products of bac-
terial origin aid in the cytolytic action of the enzyme; indeed, the con-
trary seems indicated. It is worthy of recall in this connection that the
chief cytolytic action occurs in advance of the actual invasion by the
organism and where the enzyme has passed by diffusion beyond the zone
where there is any such accumulation of bacterial products as were
present in the culture broths used above.

The cytolytic action of Bacillus carotovorus compared
with that of various other organisms.

At the beginning of this article attention was called to our con-
viction that this carrot-rot bacillus is only one of a large number of
closely related organisms capable of functioning similarly as wound para-
sites. Since our earlier publication we have continued studies upon
this class of organisms in cooperation with Messrs. H. A. Harding
and F. C. Stewart of the Geneva, New York, experiment station, and
W. J. Morse of the Vermont station. The full results of these studies
will appear soon as a joint publication from these two stations. Previous
to this publication it will not be feasible to discuss the details. We
will merely state that we have had under observation some forty strains
of organisms isolated in one or the other of these laboratories from the
rotting tissues of cabbage, turnip or Amorphophallus simlense,
and in addition five organisms from other sources, as follows: Harri-
son’s Bacillus oleraceae, a cauliflower-rot organism from Ontario ;
Spieckermann’s organism of the soft-rot of kale from Germany;
Van Hall’s two iris-rot organisms from Holland, Bacillus omni-
vor us and Pseudomonas iridis; and an organism sent us from
Krai’s laboratory, Prag, as Pseudom onas destructans, Potter’s
white soft-rot of turnip from England. This latter as we have it proves
to be a bacillus, and therefore cannot be Potters’ original organism.

Comparative trials have shown that these soft-rot organisms, although
from different vegetables and widely separated regions, are remarkably
similar in enzymic activity. The only exceptional one is Pseudomonas
iridis, and this as we have it is non-pathogenic. The others all induce
similar soft-rots of various vegetable tissues, and from cultures of each
a cytolytic enzyme has been secured indistinguishable in kind from that
produced by B. carotovorus. The only differences are minor ones
in quantity of enzyme production or degree of activity shown. In all
alike the action consists in solution of the middle lamella, and stops

270

L. R. Jones,

short of the complete solution of the cellulose layer of the wall (the
invasion of the organisms being strictly intercellular). With all alike
there is absence of diastatic (amylolytic) action.

These observations are in general accord with the records regarding
these and similar soft-rot organisms heretofore published, with the ex-
ception of Potter’s. He reports the penetration of the walls by
Pseudomonas destructans and the development of a diastatic
ferment. We would again call attention to the fact that we have not
his original organism.

While the production of this middle-lamella dissolving enzyme is
thus characteristic of a numerous class of bacterial wound parasites, we
should note in contrast that Smith1) has found complete solution of
the cell walls to occur in the black-rot of the turnip caused by Pseudo-
monas campestris. This appears, however, to be a slower action
of which the details have not been worked out. That we might compare
such complete wall solution with the partial solution shown by B. caroto-
vorus we repeated Newcombe’s experiments with Taka-diastase2).
Our results were in close agreement with those reported earlier by him,
showing a rapid solution of the inner or cellulose lamellae of parenchymatous
walls followed ultimately by a ‘ slower solvent action upon the middle
lamella or pectic layer. This action is, therefore, quite different from
that occurring with B. carotovorus.

Classification and nomenclature of cytolytic enzymes.

In connection with these studies we have been led to compare the
numerous accounts of cytolytic action in literature and to speculate as
to the evidences of relationship or difference between the enzymes causing
these. Until quite recently it was contended that cellulose fermentation
might in some cases be attributed to diastase, i. e. that the distinction
between cytolytic and amylolytic fermentation was a doubtful one 3). Our
results have contributed evidence, if such were needed, to establish the
correctness of Newcombe’s4) conclusion that where these two classes
of carbohydrate-fermentation occur together, as in germinating barley
and with Taka-diastase, it is attributable to the occurence in mixture of
two distinct enzymes. Instead of classing the wall-dissolving with the
starch-dissolving enzymes the present evidence points rather to the need
of more clearly recognizing subdivisions of the “cellulose-enzymes” or
“cytases” as they are termed. A fuller understanding of the chemistry
of the cell membranes must, of course, precede any such a subdivision
that is to be permanently satisfactory. For the present we can, however,
recognize the following well-defined groups of elements in the so-called
cellulose walls : 1) true celluloses ; 2) hemicelluloses ; 3) pectic compounds.

Omelianski5) has recently shown that even the most resistant
of the true celluloses may be dissolved by bacterial action, and he con-

1) Smith, Erwin F., The effect of black rot on turnips. (U. S. Dept. Agric.
Bureau plant industry. Bull. 29. 1903.)

2) New com be, F. C. , Cellulose enzymes. (Ann. of Bot. Bd. XIII. 1899.
p. 49—81.)

3) Cf. Griiss, J., Landw. Jahrb. Bd. XXV. 1896. p. 385; Reinitzer, F.,
Hoppe-Seyler, Zeitschr. f. physiol. Chem. Bd. XXIII. 1897. p. 175.

4) Newcombe, F. C., loc. cit.

5) Omelianski, W., Ueber die Garung der Cellulose. (Centralbl. f. Bakt.
Abt. II. Bd. VIII. 1902. p. 193. etc.)

The cytolytic enzyme produced by Bacillus carotovorus etc.

271

tends that the term “cellulose fermentation” should be applied only to
this. The commoner forms of so-called “cellulose-fermentation” really
involve only the solution of the second or third of the above groups.
Green1) stated some three years ago his conclusion from the evidence
adduced by Newcombe that the enzymes which act primarily on the
middle lamella (pectic compounds) of the wall are of a different class
from those acting upon the other portions (hemicelluloses). The logical
conclusion to date therefore, is that we must recognize three clearly
definable enzymes or enzyme-groups each capable of action primarily
upon one of the above:named groups of wall elements. If this is ac-
cepted then it will conduce to clearness is we differentiate these by
names.

The enzyme of B. carotovorus and the related soft-rot bacteria
is an example of those acting strongly upon the pectic compounds but
not capable of hydrolyzing either the hemicelluloses or the true cellu-
loses. Following the customary, terminology, pectase would be the most
appropriate name for this, but unfortunately it has long been used to
designate Fremy’s clotting enzyme. If one were to accept Cross
and Be van’s2) term “pecto-cellulose” for these pectic elements of the
wall then preference might be given to the name “pecto-cellulase” for
the enzyme acting on them. The sufficient objection to this is that the
term applies to the hypothetical compound of the non-hydrolyzable
cellulose elements with the pectic elements which pass into solution
under the influence of this enzyme. The name, if adopted, would suggest
an enzyme active upon both of these groups.

The name we would adopt is that suggested by Bourquelot and
Herissey3), pectinase. It may be objected that they originally
applied this name to the enzyme found in barley malt which so changes
pectine that it cannot thereafter be clotted by Fremy’s enzyme,
pectase.

But Bourquelot4) later showed that this same extract hydrolyzes
the pectic clot, and inferring that the action is due do the same enzyme
he extended the conception of the term. This later application of it
to the enzyme capable of hydrolyzing the pectic coagulum, including the
pectic elements of the cell wall, appears to us to justify its adoption in
this sense. A further argument for this position is that its use in this
broader sense is finding acceptance with some recent writers.

As already stated, the hydrolysis of the hemicellulose layers of the

1) Green, loc. cit. p. 105.

2) Cross, C. F. and Be van, E. J., Cellulose. London 1895.

3) Bourquelot, E. et Herissey, H., Bur l’existence dans l’orge germe d’un
ferment soluble, agissant sur la pectine. (Compt. rend. T. CXXVII. 1899. p. 191.)

4) Bourquelot, E., Sur la pectase. (Journ. pharm. et chem. T. IX. 1899.
p. 56.) After the above was written a publication was received from Beijerinck and
van Delden, on the Bacteria which are active in flax-rotting (English reprint of
paper read at Meeting of Jan. 20, 1904, of Koninklijke Akademie van Wetenschappen
te Amsterdam). In this the authors conclude that the rotting is due to a bacterial
enzyme acting on the pectose layers of the cell walls, and which is probably identical
with the enzyme we have studied. They propose the new name “pectosinase” for this
enzyme on the ground that it is not identical with the “pectinase” of Bourquelot
and Herissey (footnote, p. 7). In a personal letter of August 16, 1904, Prof.
Beijerinck states that he now thinks this is “possibly only a more concentrated so-
lution of the malt-enzyme called cytase by Brown, Morris and Escombe, and
pectinase by Bourquelot and Herissey“, although he adds “the difference is
obvious and unexplained by physical influences, etc.”

272 Jones, The cytolytic enzyme produced by Bacillus carotovorus etc.

wall was more rapid with Taka-diastase than that of the pectic layer.
We believe the evidence adduced by Newcombe must lead one to
consider this action as due to the occurrence in the Taka-diastase of a
smaller amount of the enzyme pectinase in mixture with a larger
amount of another enzyme which acts primarily on the hemicellulose
elements of the wall. Accepting this conclusion, what name should be
applied to the latter?

The one used by Oppenheimer1), cellulase, is too general, since
it implies action on all classes of cellulose. The enzyme under dis-
cussion acts only on the hemicelluloses, and therefore hemicellulase
is the preferable name for it. This is self-explanatory and leaves the
name cellulase to be applied either in a general way to all cellulose
enzymes, or, as seems preferable to us, reserves it for application to
the enzyme capable of hydrolyzing true cellulose which Omelianski
has recently shown to exist.

The tendency is now to use the words pectinase and cellulase in a
vague way as synonymous with the term cytase, i. e. as applicable to
cytohydrolytic enzymes in general. If they be restricted to the more
exact usage defined above it leaves the words cytase and cytolyst as
convenient and satisfactory terms for use in the broader sense to in-
clude in a general or indefinite way all enzymes capable of hydrolyzing
the cell walls.

1) Oppenheimer, C., Ferments, and their action. (Eng. ed.) 1901. p. 187.